Corresponding author : Hiroshi Takahashi, Department of Orthopaedic Surgery, Toho University School of Medicine, Japan, 6-11-1 Omori-nishi Otaku Tokyo, Japan, Zip-Code 143-8541, Tel: 03-3762-4151; Fax: 03-3763-7539; E-mail: firstname.lastname@example.org
Received: February 17, 2014 Accepted: March 10, 2014 Published: March 17, 2014
Citation:Takahashi H, Iida Y, Yokoyama Y, Takamatsu R, Fukutake K, et al. (2014) Recurrent Lumbar Disc Herniation after Microendoscopic Discectomy. J Spine Neurosurg 3:3 doi:10.4172/2325-9701.1000141
Introduction: The purpose of this study was to investigate the
incidence and risk factors of recurrent lumbar disc herniation (LDH)
after microendoscopic discectomy (MED).
Methods: The subjects were 210 patients who underwent MED
for LDH performed by the same operator at our hospital. There
were 132 male and 78 female patients. The treated level was L3/4
in 6 patients, L4/5 in 88, and L5/S in 116. The mean duration of
postoperative follow-up was 72.0 ± 36.4 months. The age, sex, BMI,
level of LDH, type of LDH, smoking habit, diabetes mellitus (DM),
and learning curve of the surgeon were subjected to multiple logistic
regression analysis to identify risk factors for recurrence.
Results: The recurrence rate was 8.58%, and the mean time to
recurrence was 24.18 months. Of the 18 patients, 4 required
reoperation and the remaining 14 patients received conservative
treatment. On multiple logistic regression analysis, none of the
examined factors—age, sex, BMI, level of the LDH, type of herniation,
smoking habit, past history of DM, and surgical experience of the
operator—were a significant risk factor for recurrence.
Conclusion: The recurrence rate was 8.58%, which is comparable
with that of open discectomy. Many cases of recurrence occurred
relatively early after surgery. Recurrence was seen in 6 of 18
patients within 6 months after surgery, but the mean time was about
2 years. The factors responsible for recurrence were unclear.
Discectomy for Lumbar disc herniation (LDH) is useful for
sciatica patients who do not respond to conservative treatment.
Open discectomy (OD) is the gold standard and the outcomes are
mostly favorable, although it depends on the evaluation method [1-4]. However, reoperation is not uncommon for conditions such as
recurrent herniation, new herniation at a different level, postoperative
scar, postoperative hematoma, infection, facet syndrome, secondary
spinal canal stenosis, and intervertebral instability. Recurrent
herniation is the most common reason for reoperation, and the
reported incidence is 5-11% [5-7]. Reasons for the variation in the reported recurrent herniation rates are inconsistency in the definition
of recurrent herniation, variations in treatment, and varying duration
of observation of the disease course.
Microdiscectomy (MD) became popular after establishment of
the OD, and microendoscopic discectomy (MED) and Percutaneous
endoscopic lumbar discectomy (PELD) are now becoming widespread
as non-invasive surgeries that facilitate early rehabilitation. Despite
these procedural advancements, reoperation is still unavoidable for
conditions that require reoperation, including recurrent herniation.
The aim of the present study was to investigate the incidence of
postoperative recurrent herniation and identify risk factors of
recurrence in MED-treated patients.
Materials and Methods
This study was performing according to a protocol approved by
the Institutional Review Board (IRB) of Toho University School of
Medicine. Informed consent was obtained from all patients. The subjects
were 210 patients who underwent MED for disc herniation after 2000
at our hospital. MRI was performed in all patients, and disc herniation
was diagnosed based on MRI findings in addition to clinical symptoms.
Patients with herniation at the level consistent with radicular symptoms
were selected. When hernia was absent on MRI and only narrowing of
the lateral recess was noted, the patient was diagnosed with spondylotic
radiculopathy and excluded from the study.
MED was performed by the same operator. Patients with a
past history of lumbar surgery, those undergoing reoperation for
herniation, those requiring surgery for two intervertebral segments,
and those requiring open conversion were excluded from the study.
A total of 132 male and 78 female patients were examined. The mean
age at the time of surgery was 40.3 ± 15.6 years. The treated level was
L3/4 in 6 patients, L4/5 in 88, and L5/S in 116. The type of herniation
was protrusion (type P) in 94 patients, subligamentous extrusion
(Type SE) in 68, transligamentous extrusion (Type TE) in 29, and
sequestration (Type S) in 19. The mean duration of postoperative
follow-up was 72.0 ± 36.4 months (range, 7-144 months).
No drain was inserted in 87 patients while a suction drain was
placed for 24-48 hours postoperatively in the remaining patients.
For the disc resection procedure, limited disc removal (LD) i.e.,
herniotomy was performed in all patients and aggressive removal
of the whole disc (aggressive discectomy, AD) was not required.
Patients wore a corset after surgery and were encouraged to mobilize
the following day. Exercise and heavy labor were prohibited for three
Recurrent herniation was defined as reappearance of preoperative
symptoms after the absence of symptoms for at least 1 week, and
required MRI confirmation of disc herniation at the same level. Age
[8,9], sex [8,9], BMI, level of the herniated disc, type of herniation
, smoking , past history of diabetes mellitus (DM) , and
surgical experience of the operator were subjected to multiple logistic
analysis to identify factors associated with recurrence. The surgical
experience of the operator was assessed by determining the number of
operated patients; the following 3 categories were used: 1–50, 51–100,
and > 100 patients.
The association between the response variable of recurrence and
explanatory variables was investigated. Univariate logistic analysis
was performed to identify risk factors associated with recurrence.
Multiple logistic analysis was then performed to examine seven
explanatory variables (age, sex, BMI, level of the herniated disc,
number of operated patients, Type TE, Type P) after excluding four
variables with p-values exceeding 0.9 (smoking, Type SE, Type S, and
Statistical analysis was performed using Excel Statistics 2008
(Social Survey Research Information Co., Ltd, Japan) and IBM SPSS
Statistics Version 19 (IBM JAPAN Ltd., Japan). P <0.05 was regarded
as statistically significant.
The recurrence rate was 8.58% (18 of 210 patients). Recurrence
occurred within 3 and 6 months in 5 and 1 patient, respectively,
while it occurred within 1 year in 2, after 1-2 years in 4, and after 3
years in 6 patients. The mean time to recurrence was 24.18 months
(range, 1 week-6 years). Of the 18 patients with recurrence, 4 required
reoperation while the remaining 14 patients received conservative
Analysis of the relationship between the response variable of
recurrence and explanatory variables revealed that the level of the
herniated disc had a p-value of 0.0401, which suggested an association
between this factor and recurrence (Table 1); however, it did not reach statistical significance (p = 0.0502) on univariate logistic analysis
(Table 2). Multiple logistic analysis was performed to examine seven
explanatory variables (age, sex, BMI, level of the herniated disc,
number of operated patients, Type TE and Type P) after excluding
four with p-values exceeding 0.9 (smoking, Type SE, Type S and DM).
None of the variables—age [8,9], sex [8,9], BMI, level of the herniated
disc, type of herniation , smoking , past history of DM ,
and surgical experience of the operator were significant risk factors
for recurrence (Table 3).
Table 1: Evaluation of the relationship between the response variable of
recurrence and explanatory variables.
Table 2: Univariate logistic analysis.
Table 3: Multiple logistic analysis of seven variables.
A total of five patients (2.38%, 5/210) experienced dural tears, but
repairs were not necessary because all cases were pin hole injuries.
No nerve root injury occurred during surgery. Postoperative infection
occurred in 0.95% (2 of 210 patients) of patients; one resolved
after percutaneous nucleotomy and antibiotics while the other
resolved after oral antibiotics alone. The causative bacterium was
Staphylococcus epidermidis in the former case while the causative
bacterium was unknown in the latter case because the patient refused
percutaneous nucleotomy. Postoperative symptomatic hematoma
developed in 0.48% (1 of 210 patients) of cases and required an
endoscopic reoperation. In this case, cauda equina symptoms
developed immediately after the initial surgery despite placement of
a drain, thus requiring rapid evacuation of the hematoma using an
endoscope. Poor drainage due to inappropriate drain insertion may
have caused the hematoma.
Reoperation was required for 4 of 18 cases of recurrent herniation,
1 postoperative infection, and 1 postoperative hematoma. Therefore,
the reoperation rate was 2.86% (6 of 210 patients).
MED has gained widespread acceptance as minimally invasive
surgery (MIS) for lumbar disc herniation in Japan. However, Teli et
al.  reported a high incidence of complications such as dural tear,
nerve root injury, and recurrent herniation after MED compared with
OD and MED. The authors attributed it to difficulty in judging the
depth of the surgical field because the images were two-dimensional.
The cost of MED was also noted to be higher than those of OD and MD.
Cheng et al.  reported that the incidence of recurrent herniation
was highest in PELD, followed by MED and OD, whereas the time to
recurrence was longest for the OD, followed by MED and PELD. MIS
including MED and PELD is performed based on two-dimensional
images, and excision of residual or hidden fragments is difficult
because of the limited operative field. The authors considered this as a
possible reason for the higher recurrence in MIS procedures, similar
to the reason proposed by Teli et al. . LD is frequently employed
in endoscopic discectomy, but Mc Girt et al.  reported that the
recurrence rate was higher after LD than after AD. They also reported
that postoperative lumbar and leg pain recurred more frequently
after AD after ≥2 years of follow-up. However, in a nationwide cohort
study reported by Kim et al.  in which fusion, laminectomy, open
discectomy, endoscopic discectomy and nucleolysis for lumbar disc
herniation were compared, the reoperation rates were found to be
13.8% and 12.4% for open and endoscopic discectomy, respectively.
This finding was based on evaluation of either open or endoscopic
discectomy as the initial surgery for lumbar disc herniation. Since
the reported outcomes of lumbar disc herniation treated with MIS
procedures have been inconsistent, we investigated the surgical
outcomes of patients treated in our institution. Our findings revealed
a recurrence rate of 8.58%, which is comparable with that after open
Recurrence occurred relatively early after surgery in many
cases, as was seen in other studies . Recurrence occurred within
6 months after surgery in 6 of the 18 cases, but the mean time was
about 2 years. The reoperation rate was 2.86% (6 of 210 patients),
which was lower than those in previous reports. The reoperation
rate was reduced because 14 of the 18 recurrent cases responded to
conservative treatment. In the initial surgery of patients with recurrent
herniation, the posterior longitudinal ligament had been incised or a
protrusion perforating the posterior longitudinal ligament had been
excised. Therefore, blood vessels may have invaded the recurrent
herniated tissue. Recurrent herniation may spontaneously resolve in
many cases if pain can be controlled, and not all cases require surgical
treatment. As described above, the definition of recurrent herniation
is inconsistent and the indication for surgery is variable including the
timing of surgery, and these may markedly influence the incidence of
recurrence and reoperation rate. We performed MRI in all patients
with recurrence of symptoms, which may have increased the reported
recurrence rate. If MRI had not been performed for cases in which
conservative treatment was effective and resolved the symptoms,
symptomatic cases would not be included in the determination of
recurrent herniation, which may result in a lower recurrence rate.
The age [8,9], sex [8,9], BMI, level of the herniated disc, type of
herniation , smoking habit , past history of DM , and
surgical experience of the operator were possible risk factors for
recurrence, but none were significantly associated with recurrence,
which is consistent with the finding by Hakkinen et al. .
A bright and wide visual field comparable with those in open
discectomy and MD is possible in MED because a 25° obliqueviewing
endoscope is used, the camera is present in the wound and is
not obstructed by surgical tools or the operator’s hand, and the light
source is adjustable.
Despite the two-dimensionality of the images and narrowness
of the working space in MIS, the removal of residual fragments
should be easy if the preoperative imaging is examined closely, since
the cylindrical retractor insertion angle and camera position can be
readily changed. Surgical skill in the procedure can be acquired by
performing open discectomy and receiving appropriate training.
The outcomes of endoscopic surgeries such as MED and PELD are
markedly influenced by the operator’s skill. For discectomy, AD
may be advantageous in reducing the recurrence rate, but LD may
inhibit disc degeneration to a greater degree and prevent recurrence
of lumbar and leg pains.
As postoperative pain is mild, excessive exercise and early return
to work are possible causes of recurrence after MIS in Japan . To reduce the recurrence rate, it is necessary to explain potential
complications and postoperative restriction of work and exercise to
patients before the surgery.
Several study limitations warrant mention. First, comparison
with open discectomy could not be performed because MED has
been performed as the initial surgery for disc herniation at our
hospital since 2000. Second, the indication for surgery, observation
of the clinical course, and evaluation of outcomes were done by the
same operator, so investigator bias may have resulted. However, the
intermediate-term outcomes of MED were favorable. In addition, no
risk factor for recurrence after MED was identified statistically.
Authors reported with no conflict of interest and financial support